CeylonDemon

[madushan1996]

CeylonDemon 2.0 :- released under GPL v3

Public repo :- https://github.com/Madushan996/CeylonDemon2.0
Release page :- https://github.com/Madushan996/CeylonD ... s/tag/v2.0

This is an experimental chess engine built on top of NARC engine codebase. The evaluator, Resonance, is a dual-frame NNUE original to CeylonDemon.

This engine was created as a part of my study project , explore different chess engine architectures, Training methods and Custom Training data corpuses. Because of that successive versions can be radically different from the current one.


Strength: 2719.5 ± 32.8, a locally transferred CCI-STC estimate at 10+0.1, 1 thread, 64 MB hash, over 320 games against 8 anchors.It is estimated around 3250 - 3350 on CCRL blitz scale(Note that this is a rough estimate based on CCI STC estimated rating).



Unique features of CeylonDemon 2.0.

1. Derived from NARC Engine codebase
2. Dual-frame NNUE architecture (AEGIS / LANCE)
3. Four accumulator slices
4. Sigma as an uncertainty signal :- sigma = |AEGIS − LANCE|
5. Search consumption of sigma
6. Four-slice incremental update rule
7. The .aa network format, its loader, and the embedded-network build
8. 73.2M self-play training corpus

If you are an engine tester, you are very welcome to test it. Independent results, games and feedback would be greatly appreciated. thankyou

[Tibono]

Hi,
the use of a dual NNUE point of view and triggering specific management leveraging the sigma signal seems to be an interesting and original way. Congrats for this engineering idea!
I currently lack bandwidth to do so, but shall definitely test and report.
Thanks for sharing your work,
All the best,
Tibono

[madushan1996]

Okay sure no worries! Thank you very much for your feedback!

[Sylwy]

You significantly underestimated the playing strength of your very (very) interesting chess engine—2719.5 ± 32.8 Elo points. I am running three parallel tests (two short ones: one against Gaviota-1.0X64-AVX—rated 2869 Elo in CCRL blitz—where it plays with two handicaps: *SuperGM_4mvs.abk* book versus *Gaviota.bin* & *Gaviota gtb* for endgames; the second short test is against Yukari_2025.11.1_x64, an engine rated 3542 Elo in CCRL blitz—it trails that one by 250–350 Elo points). And in the SUPREME_TOURNAMENT (featuring 16 engines that make up the world elite), it has played some beautiful games! CONGRATULATIONS!
Note: I am using the BMI2 version, which yields a better benchmark score than the AVX2 version on my test PC.











[pgn]
[Event "CeylonDemon-2.0_x64-bmi2_SHORT_TEST_2"]
[Site "ISR 3"]
[Date "2026.08.30"]
[Round "1"]
[White "CeylonDemon-2.0_x64-bmi2"]
[Black "Gaviota-v1.0-win64-AVX"]
[Result "1-0"]
[BlackElo "2200"]
[ECO "B36"]
[Opening "Sicilian"]
[Time "20:30:41"]
[Variation "Maroczy Bind, 5...d6 6.Nc3 Bg7"]
[WhiteElo "2200"]
[TimeControl "240+2"]
[Termination "adjudication"]
[PlyCount "94"]
[WhiteType "program"]
[BlackType "program"]

1. Nf3 c5 2. c4 g6 3. e4 Bg7 4. d4 cxd4 5. Nxd4 {(Nf3xd4 Nb8-c6 Bc1-e3
Ng8-f6 Nb1-c3 Nf6-g4 Qd1xg4 Nc6xd4 Qg4-d1 Nd4-e6 Qd1-d2 Qd8-a5 Ra1-c1 d7-d6
h2-h4 Bc8-d7 h4-h5 Bd7-c6 Rh1-h4 Bg7-e5 h5xg6 h7xg6 Rh4xh8+ Be5xh8 Rc1-d1)
+0.76/22 8} d6 {(d6 6.Nc3 h5 7.Be3 Nf6 8.h3 Nbd7 9.Be2 b6 10.O-O Nc5 11.Qc2
Bb7 12.f3 a5 <>>) -0.32/16 23} 6. Nc3 {(Nb1-c3 Nb8-c6 Bc1-e3 Bg7xd4 Be3xd4
Nc6xd4 Qd1xd4 Ng8-f6 Bf1-e2 Qd8-b6 Qd4xb6 a7xb6 f2-f3 Bc8-d7 a2-a3 e7-e5
O-O-O Ke8-e7 Rd1-d2 Rh8-d8 Rh1-d1 Bd7-a4 Nc3xa4 Ra8xa4 Kc1-b1 Ra4-a5)
+1.02/21 10} Nc6 {(Nc6 7.Be3 Nxd4 8.Bxd4 Bxd4 9.Qxd4 Nf6 10.Be2 O-O 11.O-O
Be6 12.Qd2 Qb6 13.b3 Qc5) -0.31/15 8} 7. Be3 {(Bc1-e3 Nc6xd4 Be3xd4 Bg7xd4
Qd1xd4 Ng8-f6 Bf1-e2 Qd8-b6 Qd4xb6 a7xb6 f2-f3 Bc8-d7 Ra1-d1 Ra8-a5 Ke1-d2
e7-e6 Kd2-e3 Ke8-e7 g2-g4 Bd7-c6 h2-h4 h7-h5 g4-g5 Nf6-d7 b2-b4) +1.02/21
6} Nxd4 {(Nxd4 8.Bxd4 Nf6 9.Be2 O-O 10.O-O b6 11.e5 dxe5 12.Bxe5 Bb7 13.b4
Nd7 14.Bxg7 Kxg7 15.Bf3) -0.39/15 8} 8. Bxd4 {(Be3xd4 Bg7xd4 Qd1xd4 Ng8-f6
Bf1-e2 O-O O-O Bc8-d7 Rf1-d1 Qd8-a5 a2-a3 Qa5-b6 Qd4xb6 a7xb6 a3-a4 Rf8-c8
b2-b3 Bd7-c6 f2-f3 Kg8-f8 Kg1-f2 Nf6-d7 b3-b4 Kf8-e8 f3-f4) +1.34/22 15}
Nf6 {(Nf6 9.Be2 O-O 10.O-O b6 11.Qc2 Bb7 12.Rad1 Qd7 13.e5 dxe5 14.Bxe5 Qe6
15.f4 Rad8 16.b4) -0.41/16 5} 9. Be2 {(Bf1-e2 O-O O-O Bc8-d7 Qd1-c2 Qd8-a5
b2-b3 Bd7-c6 Ra1-d1 Nf6-d7 Bd4xg7 Kg8xg7 Qc2-d3 a7-a6 h2-h4 Qa5-c5 Qd3-d2
Nd7-e5) +1.48/19 9} O-O {(O-O 10.O-O b6 11.Qc2 Bb7 12.Rad1 Qc7 13.Nd5 Nxd5
14.exd5 Bxd4 15.Rxd4 e5 16.dxe6 fxe6 17.Bg4 e5 18.Be6+ Kg7) -0.44/16 5} 10.
O-O {(O-O Bc8-d7 Qd1-d3 Qd8-a5 b2-b3 a7-a6 a2-a4 Bd7-c6 Ra1-b1 Qa5-b4
Nc3-d5 Bc6xd5 e4xd5 Nf6-d7 Bd4xg7 Kg8xg7 Qd3-e3 Nd7-c5) +1.71/18 6} b6 {(b6
11.Qc2 Bb7 12.Rad1 Qc7 13.Nd5 Nxd5 14.exd5 Bxd4 15.Rxd4 e5 16.dxe6 fxe6
17.Bg4 e5 <TT>) -0.44/16 8} 11. Qd2 {(Qd1-d2 Bc8-b7 f2-f3 Ra8-c8 Bd4-e3
Nf6-d7 Rf1-d1 a7-a6 Ra1-c1 Bg7-e5 b2-b3 Bb7-a8 Nc3-d5 e7-e6 f3-f4 Be5-g7
Nd5-c3 Nd7-c5 Qd2xd6 Bg7xc3) +1.78/20 10} Bb7 {(Bb7 12.Rad1 Re8 13.Qc2 Qc7
14.Nb5 Qc6 15.Bf3 Qc8 16.e5 Bxf3 17.gxf3 dxe5 18.Bxe5 Qc5 19.Bxf6 Bxf6
20.Rd5) -0.36/16 6} 12. f3 {(f2-f3 Ra8-c8 Bd4-e3 Nf6-d7 Rf1-d1 a7-a6 Ra1-c1
Bb7-a8 b2-b4 Bg7-e5 Kg1-h1 a6-a5 a2-a3 Ba8-c6 Be3-g1 h7-h5 Bg1-e3 e7-e6
f3-f4 a5xb4 a3xb4) +1.87/20 6} Nd7 {(Nd7 13.Rad1 Rc8 <>>) -0.32/15 6} 13.
Be3 {(Bd4-e3 Ra8-c8 Ra1-d1 a7-a6 Rd1-c1 Bb7-c6 Rf1-d1 Nd7-c5 b2-b3 Bc6-b7
Kg1-h1 Bb7-a8 b3-b4 Nc5-d7 Be3-g1 Nd7-e5 Nc3-d5 Ba8xd5 e4xd5) +1.90/19 9}
h5 {(h5 14.Bd4 Bxd4+ 15.Qxd4 a5 16.Nd5 Rc8 17.b3 <>>) -0.16/14 5} 14. b3
{(b2-b3 Ra8-c8 Kg1-h1 a7-a6 Ra1-c1 Bg7-e5 Rf1-d1 Bb7-a8 b3-b4 e7-e6 Nc3-a4
Ba8-c6 b4-b5 Bc6-b7 f3-f4 a6xb5 f4xe5 b5xa4 e5xd6 Bb7xe4) +2.05/20 10} Kh7
{(Kh7 15.Bd4 Bh6 16.Qc2 Ne5 17.a3 Nc6 18.Bf2 Rc8 19.Rad1 Rg8 20.Nd5 Ne5
<>>) -0.14/15 5} 15. Rad1 {(Ra1-d1 a7-a5 h2-h3 Nd7-c5 Be2-d3 Ra8-c8 Bd3-c2
Kh7-g8 Be3-h6 a5-a4 Bh6xg7 Kg8xg7 f3-f4 Rc8-c7 Kg1-h1 a4xb3 a2xb3 e7-e6)
+2.24/19 5} a5 {(a5 16.Qc2 Nc5 17.Nd5 Rc8 18.g3 <>>) -0.17/15 7} 16. Bd3
{(Be2-d3 Bb7-c6 h2-h3 Nd7-c5 Bd3-b1 Ra8-b8 Kg1-h1 Kh7-g8 Bb1-c2 Rb8-c8
a2-a3 Bc6-a8 Be3-h6 a5-a4 Bh6xg7 Kg8xg7 b3-b4 Nc5-e6 Bc2xa4 Rc8xc4 Nc3-d5)
+2.06/21 12} Nc5 {(Nc5 17.Bb1 Bc6 18.Nd5 Rb8 19.f4 <>>) -0.16/15 6} 17. Bc2
{(Bd3-c2 Bb7-c6 Kg1-h1 Ra8-c8 h2-h3 Kh7-g8 Be3-g5 Rf8-e8 Bg5-h6 e7-e6
Bh6xg7 Kg8xg7 f3-f4 Qd8-f6 Qd2-e3 Re8-d8 f4-f5 Qf6-e5 f5xg6 f7xg6) +2.22/19
5} Bc6 {(Bc6 18.Bd4 Rb8 19.Bxg7 Kxg7 20.Qd4+ f6 21.Qe3 b5 <>>) -0.26/14 6}
18. Rf2 {(Rf1-f2 Ra8-c8 h2-h3 Qd8-d7 Be3-d4 Bg7xd4 Qd2xd4 Kh7-g8 f3-f4
Qd7-e6 Rf2-e2 h5-h4 Re2-e1 Bc6-b7 Kg1-h1 Bb7-a8 Nc3-b5 Rf8-d8 f4-f5)
+2.00/19 14} Rg8 {(Rg8 19.Bd4 Bh6 20.Qe1 Ne6 21.Be3 Bg7 22.Nd5 Nc5 23.Re2
e6 24.Nc3 Qe7 25.Nb5) -0.17/14 8} 19. h3 {(h2-h3 Rg8-f8 a2-a3 Ra8-c8 Kg1-h1
Bc6-b7 Nc3-b5 Kh7-g8 b3-b4 Nc5-e6 Bc2-b3 Rc8-a8 Qd2-c2 Bg7-f6 f3-f4 a5xb4
a3xb4) +2.37/18 13} Rc8 {(Rc8 20.Rff1 Bh8 21.a3 <>>) -0.13/14 11} 20. Bd4
{(Be3-d4 Qd8-f8 f3-f4 Bg7xd4 Qd2xd4 Qf8-g7 Qd4-e3 Rg8-d8 f4-f5 e7-e6 Rd1-f1
Rd8-f8 f5-f6 Qg7-h6 Qe3-g3 h5-h4 Qg3xd6 Qh6-e3 Qd6-d2 Qe3xd2 Rf2xd2)
+2.49/18 6} Bh6 {(Bh6 21.Be3 Bg7 22.Rff1 <TT>) -0.13/14 6} 21. f4 {(f3-f4
e7-e6 Bd4-e3 Qd8-e7 a2-a3 Bh6-g7 Be3-d4 e6-e5 f4xe5 d6xe5 Bd4-e3 Rg8-d8
Nc3-d5 Bc6xd5 c4xd5 Rd8-d6 b3-b4 a5xb4) +1.96/19 5} a4 {(a4 22.Nxa4 Bxa4
23.bxa4 Ne6 24.Qb4 Qc7 25.Be3 Nc5 26.Rd5 e6 27.Rd1 Bg7) -0.25/13 6} 22. b4
{(b3-b4 Nc5-d7 Qd2-e2 e7-e6 Nc3xa4 Qd8-c7 Na4-c3 Rg8-f8 Bd4-e3 Bh6-g7
Nc3-b5 Bc6xb5 c4xb5 e6-e5 f4-f5 Bg7-h6 Be3xh6 Kh7xh6 Qe2-d2+ Kh6-g7 Qd2xd6
Nd7-f6) +4.97/19 4} Nd7 {(Nd7 23.Qe2 e6 <>>) -1.56/14 4} 23. Qe2 {(Qd2-e2
e7-e6 Nc3xa4 Qd8-c7 Na4-c3 Rg8-f8 Bd4-e3 Bh6-g7 Qe2-d2 Kh7-g8 a2-a4 Rf8-e8
Nc3-b5 Bc6xb5 c4xb5 d6-d5 e4-e5 Qc7-c3 g2-g4 h5xg4) +4.80/20 4} e6 {(e6
24.e5 dxe5 25.fxe5 Qe7 26.Nxa4 Rb8 27.b5 Ba8 28.Qxh5 Kg7 29.Qe2 Rgd8
30.Rff1 Qa3) -1.72/15 7} 24. Nxa4 {(Nc3xa4 Qd8-c7 Na4-c3 Rg8-f8 Bd4-e3
Bh6-g7 Qe2-d3 Kh7-g8 a2-a4 Rf8-d8 Rf2-d2 Nd7-f6 Be3-f2 Bc6-b7 Bc2-b3 Rd8-d7
Bf2-h4 Qc7-c6 f4-f5 g6xf5) +5.14/20 6} Ra8 {(Ra8 25.b5 Bb7 26.e5 d5 27.Qe3
Qe7 28.cxd5 Bxd5 29.Bxb6 Bc4 30.Nc3 Nxb6 31.Qxb6 Ra3) -1.47/14 6} 25. b5
{(b4-b5 Bc6-b7 e4-e5 d6-d5 c4xd5 Bb7xd5 Na4-c3 Bd5-b7 Bd4xb6 Qd8-e7 Bb6-e3
Rg8-c8 Qe2-d2 Nd7-c5 a2-a4 Kh7-g8 Qd2-d6 Qe7xd6 Rd1xd6 Nc5-d7 Rd6xd7 Rc8xc3
Be3-d4) +6.84/18 4} Bb7 {(Bb7 26.e5 d5 27.cxd5 Bxd5 28.Nc3 Bb7 29.Bxb6 Qc8
30.Bd4 Ra3 31.Qxh5 Qc7 32.Qh4 Re8) -2.43/15 1} 26. e5 {(e4-e5 d6-d5 c4xd5
Bb7xd5 Na4-c3 Bd5-b7 Qe2xh5 Qd8-e7 Qh5-g4 Ra8-c8 h3-h4 f7-f5 e5xf6/ep
Nd7xf6 Bd4xf6 Qe7xf6 Nc3-e4 Bb7xe4 Bc2xe4 Rg8-d8 Rd1xd8 Rc8xd8) +6.70/19 4}
d5 {(d5 27.cxd5 Bxd5 28.Nc3 Bb7 29.Bxb6 Qc8 30.Bd4 Qc7 <>>) -2.69/16 14}
27. cxd5 {(c4xd5 Bb7xd5 Na4-c3 Bd5-b7 Qe2xh5 Qd8-e7 Qh5-g4 Ra8-a3 Bc2-e4
Nd7-c5 Be4xb7 Qe7xb7 Qg4-h4 Rg8-a8 Rf2-e2 Ra8-a5 Rd1-d2 Qb7-a8 Re2-e3
Qa8-h8 Rd2-d1 Ra5-a8 Re3-g3) +6.65/20 4} Bxd5 {(Bxd5 28.Nc3 Bb7 29.Bxb6 Qe7
30.Bc7 Rgc8 31.Bd6 Qd8 <>>) -2.82/16 4} 28. Nc3 {(Na4-c3 Bd5-b7 Bd4xb6
Qd8-e7 Bb6-e3 Rg8-c8 Nc3-a4 Bb7-d5 Bc2-b3 Qe7-a3 Rd1-d3 Ra8xa4 Bb3xd5
Qa3-e7 Bd5-b3 Ra4-e4 g2-g4 h5xg4 h3xg4 Re4-b4 Rd3-d6 Nd7-c5 Bb3-c2 Qe7-a7
Rf2-f1) +7.25/20 4} Bb7 {(Bb7 29.Bxb6 Qe7 30.Bc7 Rgc8 31.Bd6 Qh4 32.Qe3 Rc4
33.Rd4 Rac8 34.Rxc4 Rxc4 35.Ne4 Qd8 36.Bd3) -2.83/16 3} 29. Qxh5 {(Qe2xh5
Ra8-a3 Qh5-g4 Qd8-e7 Bc2-e4 Nd7-c5 Be4-c6 Rg8-d8 Rf2-f3 Ra3-a7 h3-h4 Bb7xc6
b5xc6 Ra7-c7 h4-h5 Rd8-g8 Rf3-h3 Rc7xc6 h5xg6+ Rg8xg6 Qg4-f3 Rc6-c8 Nc3-b5
Qe7-b7 Qf3xb7 Nc5xb7 Bd4xb6 Rc8-c4 Rd1-d7) +6.94/22 8} Ra3 {(Ra3 30.Qe2 Qc7
31.Bb3 Rd8 32.Ne4 Bxe4 33.Qxe4 Nc5 34.Qc6 Qxc6 35.bxc6 Nxb3 36.axb3 <TT>)
-1.89/15 4} 30. Qg4 {(Qh5-g4 Qd8-e7 Nc3-e4 Bb7xe4 Bc2xe4 Nd7-c5 Be4-c6
Nc5-a4 g2-g3 Qe7-b4 Kg1-g2 Rg8-c8 h3-h4 Na4-c3 Bd4xc3 Qb4xc3 Kg2-h3 Kh7-g7
Rf2-f3 Qc3-b4 Rf3xa3 Qb4xa3 h4-h5) +6.30/21 14} Qc7 {(Qc7 31.Be4 Ba8 <>>)
-2.21/14 5} 31. h4 {(h3-h4 Nd7-f8 Nc3-e4 Bb7xe4 Bc2xe4 Qc7-a7 Be4-c6 Rg8-h8
g2-g3 Bh6-g7 Rf2-d2 Qa7-a4 Kg1-h1 Kh7-g8 Kh1-h2 Qa4-c4 Bd4xb6) +8.64/18 3}
Rg7 {(Rg7 32.h5 Nf8 33.Ne4 Rxa2 34.Bb1 Rxf2 35.Nf6+ Kh8 36.Kxf2 Qc4 37.hxg6
Rxg6 38.Bxg6 Nxg6 39.Qh3 Qc2+ 40.Ke1) -2.62/14 3} 32. h5 {(h4-h5 Nd7-f8
Qg4-h3 Rg7-g8 h5xg6+ f7xg6 Bc2-b3 Rg8-h8 Bb3xe6 Nf8xe6 Qh3xe6 Rh8-f8 Qe6-h3
Bb7-e4 g2-g4 Rf8-d8 Rf2-h2 Kh7-g8 Qh3xh6 Ra3xc3) +9.82/18 6} Nf8 {(Nf8
33.Ne4 Kh8 34.Nf6 Rxa2 35.Ne8 <TT>) -2.94/15 3} 33. Qh3 {(Qg4-h3 Qc7-c8
h5xg6+ f7xg6 Bc2-b3 Rg7-d7 Rf2-d2 Kh7-g7 Qh3-e3 Rd7-f7 g2-g3 Kg7-g8 Nc3-e4
Bb7xe4 Qe3xe4 g6-g5 Rd2-c2 Qc8-e8 Rc2-h2 Rf7-h7 f4xg5 Bh6xg5 Rh2xh7 Nf8xh7)
+10.19/18 5} Rg8 {(Rg8 34.hxg6+ fxg6 35.Bb3 Kg7 36.Na4 Rh8 37.Bxb6 Qf7
38.Bxe6 Nxe6 39.Qxa3 Bxf4 40.Nc5 Bh2+ 41.Kf1) -4.14/14 5} 34. hxg6+
{(h5xg6+ f7xg6 Bc2-b3 Kh7-g7 Qh3-e3 Qc7-e7 Bd4xb6 Qe7-h4 Qe3-h3 Qh4xh3
g2xh3 Ra3-a8 Rd1-d6 Ra8-c8 Nc3-e2 Bb7-e4 a2-a4 Rg8-h8 Ne2-d4) +10.62/19 6}
fxg6 {(fxg6 35.Bb3 Ra8 <>>) -4.67/15 3} 35. Bb3 {(Bc2-b3 Rg8-h8 Qh3-e3
Nf8-d7 Nc3-e4 Bb7xe4 Qe3xe4 Bh6-f8 Qe4-c6 Qc7xc6 b5xc6 Nd7-c5 c6-c7 Bf8-e7
g2-g4 Kh7-g7 Bd4xc5 Be7xc5 Rd1-d7+ Kg7-h6 g4-g5+ Kh6-h5 Rd7-h7+ Rh8xh7)
+10.62/18 4} Rxb3 {(Rxb3 36.axb3 Rh8 37.Na4 Qd8 38.Nxb6 Kg8 39.Qg3 Bg7
40.Nc4 Bd5 41.b6 Qh4 42.Qxh4 Rxh4 43.Na5 Nd7) -4.91/16 11} 36. axb3 {(a2xb3
Rg8-h8 Qh3-e3 Kh7-g8 Bd4xb6 Qc7-h7 Qe3-h3 Qh7-f7 Qh3-g4 Nf8-d7 Bb6-e3
Bh6-f8 Rf2-d2 Bb7-d5 Nc3xd5 Nd7xe5 f4xe5 Rh8-h1+ Kg1xh1 Qf7-h7+ Kh1-g1)
+10.86/20 3} Rh8 {(Rh8 37.Na4 Qd8 38.Nxb6 Kg8 39.Qg4 Qe7 40.Nc4 Bg7 41.Nd6
Rh4 42.Qg5 Qxg5 43.fxg5 Bd5 44.b6 Bxb3) -5.07/16 3} 37. Qe3 {(Qh3-e3 Kh7-g8
Bd4xb6 Qc7-h7 Qe3-h3 Qh7-f7 Qh3-g4 Nf8-d7 Bb6-e3 Bh6-g7 Rf2-d2 Nd7-f8
Rd2-d8 Qf7-c7 Qg4xe6+ Kg8-h7 Qe6-c4 Qc7xc4 b3xc4 Kh7-g8 e5-e6) +10.86/22 5}
Nd7 {(Nd7 38.Qh3 Nf8 39.Na4 <TT>) -5.07/17 3} 38. Ne4 {(Nc3-e4 Bb7-d5
Ne4-d6 Rh8-f8 Rd1-c1 Qc7-d8 g2-g3 Kh7-g7 Rc1-c8 Qd8xc8 Nd6xc8 Rf8xc8 Kg1-f1
Kg7-h7 Rf2-h2) +11.03/16 5} Bd5 {(Bd5 39.Rc1 Qb7 40.Nf6+ Nxf6 41.exf6 Rd8
<>>) -6.00/15 7} 39. Nd6 {(Ne4-d6 Rh8-f8 g2-g3 Kh7-g7 Rd1-c1 Qc7-a7 Qe3-c3
g6-g5 f4-f5 Nd7-c5 b3-b4 g5-g4 b4xc5 Bh6xc1 Qc3xc1) +11.08/16 4} Nc5 {(Nc5
40.Rc1 Qb8 <>>) -5.68/15 4} 40. b4 {(b3-b4 Qc7-e7 b4xc5 Kh7-g8 Nd6-e4
Bh6-g7 Ne4-g5 Bg7-f8 c5xb6 Bf8-g7 Rd1-c1 Qe7-d7 b6-b7 Bg7-f8 b7-b8Q Rh8-h1+
Kg1xh1 Qd7-h7+ Kh1-g1) +11.13/17 6} Na4 {(Na4 41.Rc1 Qe7 42.Rfc2 Qh4
43.Rc7+ Bg7 44.Ne4 Bxe4 45.Qxe4 Qg3 46.Bf2 Qb3 47.R1c6 Kg8 48.Rc8+ Kh7
49.Rxh8+ Kxh8 50.Qxg6 Qxb4) -6.38/17 4} 41. Ra1 {(Rd1-a1 Qc7-d7 Bd4xb6
Rh8-a8 Bb6-d4 Kh7-g8 Qe3-g3 Kg8-h7 b5-b6 Qd7-c6 Qg3-h3 Na4-c3 Ra1xa8
Nc3-e2+ Rf2xe2) +11.13/17 4} Qa7 {(Qa7 42.Qh3 Kg8 43.Qa3 Qh7 44.Rc2 <>>)
-6.74/15 3} 42. Rc1 {(Ra1-c1 Bh6-f8 Nd6-e8 Bf8-c5 b4xc5 Kh7-g8 Ne8-f6+
Kg8-g7 Nf6xd5 Na4-c3 Bd4xc3 Kg7-g8 Qe3-h3 Kg8-g7 Rf2-f1) +11.13/16 3} Qe7
{(Qe7 43.Rfc2 Qh4 <TT>) -6.38/15 3} 43. Qg3 {(Qe3-g3 Rh8-f8 Bd4-e3 Na4-c3
f4-f5 Nc3-e2+ Rf2xe2 g6-g5 f5xe6 Kh7-g8 Nd6-f7 Kg8-h7 Be3-f4 g5xf4 Qg3-d3+
Kh7-g7) +11.40/17 7} Ra8 {(Ra8 44.Qh3 Rf8 <>>) -8.31/15 8} 44. Rfc2
{(Rf2-c2 Ra8-d8 Qg3-h3 Na4-c5 b4xc5 Qe7-f8 Bd4-e3 Kh7-g8 Rc2-c3 b6xc5
Qh3-g4 Qf8-g7 Be3xc5 Qg7-h7 Nd6-f7 Kg8xf7) +11.78/15 7} Rd8 {(Rd8 45.Rc7
Rd7 46.Rxd7 Qxd7 47.Qa3 Qe7 48.Qxa4 Bxf4 49.Rc8 Bh2+ 50.Kf1 Bb7 51.Nxb7
Qf7+ 52.Bf2 Qxb7 53.Qa8) -8.00/17 3} 45. Qh3 {(Qg3-h3 Na4-b2 Rc2-c7 Rd8-d7
Rc7xd7 Qe7xd7 Bd4xb2 Bd5-a2 Kg1-h1 Qd7-g7 Nd6-e8 Qg7-f7 Rc1-c8 Kh7-g8)
+11.93/16 6} Rd7 {(Rd7 46.g4 Qf8 47.Rc7 Qe7 48.Rxd7 Qxd7 49.g5 Qg7 50.gxh6
Qe7 51.Rc8 Qd7 52.Qg4 Kxh6 53.Qh4+ Kg7) -10.40/15 5} 46. Rc8 {(Rc2-c8
Bd5-b7 Rc8-b8 Na4-c5 b4xc5 Rd7-c7 Rb8-h8+ Kh7xh8 Qh3xh6+ Kh8-g8 Nd6xb7
Qe7-g7 Qh6xg7+ Kg8xg7 Nb7-d6 Kg7-h7 Nd6-e8 Kh7-g8 Ne8xc7 Kg8-h7 Nc7xe6)
+12.14/16 3} Bb7 {(Bb7 47.Nxb7 Qg7 48.Nd8 Nc5 <>>) -16.07/16 9} 47. Rb8
{(Rc8-b8 Na4-c5 b4xc5 Rd7-c7 Rb8-h8+ Kh7xh8 Qh3xh6+ Kh8-g8 Nd6xb7 Qe7-g7
Qh6xg7+ Kg8xg7 Nb7-d6 Kg7-g8 c5xb6 Rc7-c2 Kg1-h2 Kg8-h7 Rc1xc2) +12.16/17
7} Nc3 {(Nc5 48.bxc5 bxc5 49.Bxc5 Rc7 <>>) -17.57/16 11 Arena
Adjudication} 1-0
[/pgn]

======================================================
More: viewtopic.php?p=996965#p996965
======================================================

[Sylwy]

I just saw your estimate on the CCRL blitz scale from the first post: "It is estimated around 3250 - 3350 on CCRL blitz scale." Indeed, my test with Yukari_2025.11.1_x64 lands right around there!

[madushan1996]

I just saw your estimate on the CCRL blitz scale from the first post: "It is estimated around 3250 - 3350 on CCRL blitz scale." Indeed, my test with Yukari_2025.11.1_x64 lands right around there!

Hi thank you very much for testing my engine and amazing feedback!. I tested my engine in 10s + 0.1 s time control(So I can get more number of games quickly to minimize statistical error) . For that time control most comparable ratings that I can found was STC ratings of Computer chess Index. Those ratings are around 500 points lower than the CCRL blitz ratings of most of the engines in this catergory.

[Sylwy]

I just saw your estimate on the CCRL blitz scale from the first post: "It is estimated around 3250 - 3350 on CCRL blitz scale." Indeed, my test with Yukari_2025.11.1_x64 lands right around there!

Hi thank you very much for testing my engine and amazing feedback!. I tested my engine in 10s + 0.1 s time control(So I can get more number of games quickly to minimize statistical error) . For that time control most comparable ratings that I can found was STC ratings of Computer chess Index. Those ratings are around 500 points lower than the CCRL blitz ratings of most of the engines in this catergory.

viewtopic.php?p=997056#p997056